This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Paul Everill, Davit Lomitashvili, Nikoloz Murgulia, Ian Colvin, Besik Lortkipanidze

FORTY SEASONS OF EXCAVATION: NOKALAKEVI-TSIKHEGOJI-ARCHAEOPOLIS

Abstract. The ruins in the small village of Nokalakevi in Samegrelo, west Georgia, have attracted schol- arly interest since the first half of the 19th century. They were first excavated in 1930, confirming their identification as the remains of the fortress of Archaeopolis mentioned in early Byzantine historical sources, and known as Tsikhegoji or ‘the triple-walled fortress’ by the Georgian chroni- clers. The 40th season of excavation took place in 2015, part of an on-going collaboration be- tween the Anglo-Georgian Expedition to Nokalakevi, established in 2001, and the S. Janashia Museum expedition to Nokalakevi, which started work on the site in 1973. The fortifications en- close a naturally defensible area of approximately 20ha, with a steep limestone river gorge to the north, west and (to a lesser extent) the south, and a hilltop citadel standing more than 200m above the lower town. The site has seen human activity since at least the 8th century BC, with indications of a much earlier presence in the area. This paper seeks to outline the key results of the 40 seasons of excavation, against the backdrop of the shifting political landscape of Georgia.

Introduction. In 2015 the multi-period site of Nokalakevi in western Georgia hosted its 40th season of exca- vation. Situated in Samegrelo (Figure 1), 15km from the modern regional capital of Senaki, the ancient settlement was an important administrative and/ or military centre of Colchis and its successor states from the 8th century BC to the 8th century AD. Most famous today for the stand- ing remains of the early Byzantine period fortress of Archaeopolis, the site would have com- manded an important crossing point of the river Tekhuri (Figures 2 and 3), at the junction with a valuable strategic route that still winds through the neighbouring hills to Chkhorotsqu in central Samegrelo. Nokalakevi-Archaeopolis played a pivotal part in the major wars fought between the Byzantines and Sasanians in the South Caucasus during the sixth century AD. It was one of the key fortresses guarding Lazika (modern west Georgia) from Sasanian Persian and Iberian (East Georgian/ Kartlian) attack, and was part of a complex chain of forts and towers established along the northeastern frontier of the Byzantine Empire [Murgulia 2013; Colvin et al 2014]. During the war of AD 540-562, the Persians’ failure to take Nokalakevi-Archaeopolis from the Byzantines and their Laz allies eventually cost them control of Lazika. Semi-mythical accounts attribute the earliest fortification of the site to the Hellenistic-period, West Georgian ruler Kuji, from whom the site derives the Georgian name, Tsikhegoji (“the fortress of Kuji”). However, the earliest surviving fortifications at Nokalakevi date to the 4th century AD. They were strengthened in the 5th century, and significant additional fortifications were added in the 6th century AD, including a remodelling of defensive works around the eastern gate. The early Byzantine defensive fortifications of Nokalakevi-Archaeopolis are augmented by its topographic position (Figures 2 and 3) next to the river Tekhuri, which, to the west of the fortress, has carved a gorge through the limestone geology. Furthermore, the steep and rugged terrain to the north of the site made the citadel that was situated there largely unassailable. A wall connected this

146 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

‘upper town’ to the ‘lower’ town below, meaning that a total area of approximately 20ha is con- tained within the fortifications. Occupation appears to have been focussed in the lower town, however, on relatively level ground between the steep slope of the Unagira ridge and the gorge of the Tekhuri (Figure 2 and 3). Excavations in the area of the lower town have revealed substantial stone buildings of the 4th to 6th century AD (Figure 4), including the extant Forty Martyrs’ Church – first built in the 6th century – and two phases of a 5th century church immediately to its south, visible today only as the foundations exposed by archaeological investigation. Immediately south of these founda- tions lies the ruin of a large stone-built building, interpreted as a palace, constructed around the beginning of the 6th century AD and converted into a wine-cellar in the 16th-17th centuries. A small vaulted, stone gatehouse, or bell-tower, lies approximately 25m east of, and centred on, the first church. The surviving walls along which it is located suggest that it was constructed as the entrance to the earliest ecclesiastical precinct. Other stone structures revealed over many years’ work in the lower town include: the remains of baths along the inside of the southern for- tifications; a tunnel down to the river at the south-west of the site; a small bathhouse 35m east of the tunnel, apparently supplied by a cistern constructed up the slope, 50m to its northwest; and a rectangular building near the southeast of the area, which possibly housed the military commanders of the Byzantine and Laz garrison of the 5th/6th century. Excavations have also shed more light on the fortification works. This includes the protruding towers of the first wall, dated to the 4th century AD, made flush in the construction of the second phase in the th5 century, and the 6th century re-modelling that included moving the gate and changing the approach to it so as to prevent a frontal attack. Excavations in the upper town, or citadel, have revealed multi- phase towers at the northwest and northeast of the fortifications, and another small gate and probable guardhouse in the southwest corner (Figure 4). Beneath the early Byzantine period structures and layers of the lower town is evidence of several earlier phases of occupation and abandonment from the 8th to 1st centuries BC, which includes a substantial Hellenistic period (4th to 1st centuries BC) settlement and necropolis. Cur- rent evidence from OSL dating of ceramics indicates a prehistoric origin for settlement at the site, which is hardly surprising given the wealth of resources and easily-defendable character of the topography. Whatever the early origins of settlement at Nokalakevi, by the 8th/7th century BC there was clearly a significant population engaged in complex ritual activity unique to the region. Double-headed zoomorphic figurines (Figure 5) dating to this period have only been found at Nokalakevi and at Vani, 40km to the southeast in Imereti. These finds indicate a unified socio-cultural system that spanned the Colchian plain. The settlement appears to have become more substantial in the Early Antique period (6th/5th centuries BC), continuing to grow in the Hel- lenistic period, before the Laz kings and their Byzantine allies built the mighty fortifications at Nokalakevi that can still be seen today. The Swiss philologist Dubois de Montpéreux (1839), was the first to associate the walls of Nokalakevi with the Byzantine period fortress of Archaeopolis – though he also erroneously con- cluded that it had been built on the ruins of the Colchian city of Aea – and his sketch of the Forty Martyrs’ Church, surrounded by the decaying walls of the lower town (Figure 6), is the earliest known image of the site. Madame Carla Serena was among the western travellers attracted to the ruins, most likely, by Dubois de Montpéreux’s account and her descriptions of the area in

147 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

the mid-1870s are particularly informative. She describes the difficulty of getting to Nokalakevi in the years before she visited, as a result of the marshes by then made passable by a new road from Senaki; and the often fast-flowing Tekhuri river, being bridged at the time of her visit. Hav- ing reached the site she wrote: “As for other quite numerous ruins to be seen inside the present city, they are, it would seem, the remains of houses and churches. At the central point of the hill arises a spacious gateway, the only one that affords entry to the city, the dressed stones of which it is constructed are so massive that you ask yourself how, without the aid of machinery, these gigantic blocks could be transported.” [Serena 2015: 22]. Serena concludes her description by writing, “The thickly wooded mountain, teeming with game, which dominates these ruins is known as Mount Unagira. An excellent lunch, washed down with champagne, was offered to me at the foot of these venerable ruins, the philosophic visions of the past were thus agreeably combined with the legitimate enjoyment of the present moment.” [Serena 2015: 22]. The first archaeological excavations in Nokalakevi took place from the end of November 1930 to the end of January 1931, and were funded by the Emergency Association of German Science (Der Notgemeinschaft der Deutschen Wissenschaft) in collaboration with the National Education Commission for the Georgian Soviet Socialist Republic and the Museum of Georgia in Tbilisi [Kirchhoff 2003: 338]. The museum had been founded, in 1852, as the Museum of the Caucasian Department of the Geographic Society. It became the Caucasian Museum in 1865 and then, in 1919, during Georgia’s three year independence between the overthrow of Tsarist rule and the Soviet occupation, the Museum of Georgia. It was renamed once more in 1947 in tribute to the Georgian historian S. Janashia. In 2004 the S. Janashia Museum of Georgia, by then an important tourist attraction in Tbilisi and a curatorial organisation in its own right, became a constituent el- ement of the new Georgian National Museum, as part of the modernising of cultural heritage or- ganisations after the bloodless Rose Revolution. Throughout the 40 seasons of excavation since 1930 the museum (as the Museum of Georgia, the Janashia Museum, or the Georgian National Museum) has been one of the few constants. This paper is the first to summarise the results of those 40 seasons and, in doing so, considers them in their shifting socio-political context. Season One: The First Nokalakevi Expedition. The plans for the trial excavations in 1930, a collaboration between Weimar German and Soviet Georgian specialists, were conceived by Joseph Sauer of the Deutsches Archäologisches Institut following his visit to the region in 1929. Sauer’s friendship with Friedrich Schmidt-Ott of the Notgemeinschaft proved exceptionally useful and funding was provided for Sauer’s stu- dent, Alfons Maria Schneider, to undertake the work [Arnold 1999] with the support of Georgian specialists Levan Muskhelishvili and Giorgi Gozalishvili. Gozalishvili [1981: 243-248] noted that Giorgi Chubinashvili and Shalva Nutsubidze had presented an exhibition in Germany at the be- ginning of November 1930, and accompanied Schneider to Georgia on their return. Schneider and Gozalishvili arrived in Senaki on the 24th November, where they met local authorities before moving on to Nokalakevi. After six weeks’ of excavation, excluding days lost to winter weather [Muskhelishvili 1987: 292], the excavation was concluded by 27th January 1931 when they gave a presentation to local authorities in Senaki, before leaving for Zugdidi the following day. During this first season, the team of archaeologists and workmen traced the line of the fortifi-

148 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

cation walls, and excavated a number of towers and, to the east of the Forty Martyrs Church, an area Schneider interpreted as the ancient ‘agora’ – now believed to be the walls and bell tower of the 5th/ 6th century ecclesiastical complex. They appear to have excavated 22 test pits during their field season, though the location of them is not precisely recorded in the notebooks kept by the excavators. Research is currently being undertaken in the original Georgian and German archives to shed more light on the work of this expedition. According to the only published work to be produced at the time, Schneider concluded from the, rather slim, archaeological evidence that there is no evidence available at all that the place was already inhabited in pre-Roman times. The earliest which demonstrably survives is a little settlement on the western slope of the acropolis, which must be ascribed to sometime after the early Roman Iron Age, based on meagre finds of potsherds. [Schneider 1931: 354] He argued that in the 4th century AD a large city with strong fortifications sprang up on the site, the Tsikhegoji of the Georgian chronicles [Kartlis tskhovreba]. These original fortification walls were, in his opinion, destroyed by an earthquake at the end of the 6th century AD and re- built at the end of the 6th or the beginning of the 7th century. According to Schneider, these new walls were damaged by another earthquake and repaired with cobbles, though he was unaware that there were actually three walls. Schneider believed that the cultural layer contained 4th to 8th century AD pottery. Furthermore, a hoard of 23 gold coins (Figure 7) of the Emperor Maurice (582-602 AD) was found in one of the towers. Schneider’s expedition was the first to excavate a burial at Nokalakevi, which was found to the north of the Forty Martyrs Church beyond the precinct wall. It contained two small ceramic jugs, two bronze bracelets, five beads of glass-like paste and an irregularly shaped piece of bronze sheet. Schneider dated this burial to the 2nd or 3rd century AD, though the description of the grave goods certainly sound far more typical of those of the Hellenistic period that have been excavated since. Seasons Two to Twenty One: The Nokalakevi Expedition under the Georgian SSR. There was no further archaeological excavation at Nokalakevi for a generation, though af- ter the Second World War the standing remains were occasionally the subject of survey work or architectural analysis. The second season of excavation was not until 1968, with a handful of testpits opened by the West Georgian Exploratory Archaeological Expedition, directed by G. Grigolia [Grigolia et al 1973]. Grigolia’s team returned in 1971 and, in the process of excavating further test pits, unearthed a Hellenistic period pot burial with an associated bronze bracelet [Grigolia et al 1972]. This was the first burial to be discovered in a large ceramic vessel, a pithos, at Nokalakevi. A second was found in 1974, and six more between 1975 and 1977. Further examples were excavated between 1978 and 1988, however the majority of graves in the vast Hellenistic period necropolis, later dissected by the eastern fortifications, were flexed in humations. The burial ground was interpreted, according to Gvinchidze [1988: 25], as representing three chrono- logical phases: 1. Inhumations containing tightly flexed burials, dating tothe end of the th4 to the middle of the 3rd century BC; 2. Pithos (large ceramic vessels) graves, dating to the middle of the 3rdto the end of the 2nd century BC; 3. Inhumations containing moderately flexed burials, dating tothe end of the nd2 to the end of the 1st century BC;

149 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Twenty further Hellenistic burials were found between 2003 and 2010, and while none were within a pithos they did includecremation burials; jar/ dergi burials; and amphora burials, none of which had previously been identifiedin Nokalakevi. Concerted efforts to study and conserve the site (Figure 8) began with the fourth season of excavation in 1973, when the Department of Medieval Archaeology of the S. Janashia Museum of Georgia in Tbilisi established a Nokalakevi Expedition, led by P. Zakaraia and N. Lomouri. From 1973 until 1990 (Seasons Four to Twenty One of work at Nokalakevi) the annual excavations were well funded by the relevant agencies of the Georgian SSR. The budget from the Academy of Sci- ences increased from 5000 Roubles for excavation in 1973 to 75000 Roubles by 1990; and from 40000 Roubles (from the Department of Monument Preservation of the Ministry of Culture) for restoration and conservation of the standing remains in 1974, to 200000 Roubles by 1990. This significant sum, totalling 275000 Roubles (approximately 200000 US$) a year by 1990, enabled six-month-long field seasons of excavation and conservation. The results of the excava- tions from 1973 to 1989 were published in three volumes in the 1980s and 90s [Zakaraia 1981; 1987; 1993]. In 1973 and 74 work, reported in Zakaraia (1981), was focused on the initial assessment of the archaeological remains, which revealed a large bathhouse along the southern wall of the fortifi- cations and the south end of a large stone-built building, interpreted as a palace (Figure 9). Ex- cavations revealed that the latter had been constructed around the beginning of the 6th century AD and was converted into a wine-cellar in the 16th-17th centuries. The east gate was investigated and it was confirmed that the city had not only a ‘land’ gate but a ‘river’ gate as well. The former was the arched gateway in the east wall of the city, while the latter was actually a tunnel running down to the river at the western end of the lower town. In 1975 and 76 the excavation of the eastern fortifications continued, and this revealed that there were actually three parallel defensive walls [Zakaraia 1981]. The inner wall was constructed of dressed limestone blocks and was preserved in places to a height of up to 4 metres. This wall was reinforced by square protruding towers, of which two protected the gate house and one situated half way up the hill slope protected the lower fortifications from being overtopped. A second wall was subsequently constructed between, and flush with, these protruding towers. Later excavations, in the 1990s and early 2000s, revealed that the first wall has a very substantial foundation; however the second wall lacks any proper foundations, its wide base resting close to the surface and its upper levels tapering. The third wall, of large ashlar blocks, was clearly added last. The excavators dated the first wall to the th4 century AD, the second to the turn of the 4th and 5th centuries, and the third to the end of the 5th or the early 6th century. At the end of 1974 work was begun on the construction of a building for the management of the Nokalakevi State Farm, which now houses the local museum. This was situated 100 metres to the east of the lower terrace fortifications. An archaeological watching brief on the work re- corded a pot burial at the point of convergence of the trenches dug for the foundations of the north and the east walls [Zakaraia 1981]. Further excavations in this area in the following years, up to 1977, revealed 24 more graves belonging to an ancient cemetery. Of these, two inhuma- tions date back to the 5th or the middle of the 3rd centuries BC; six pot burials to the middle of the 3rd or to the 2nd century BC [Zakaraia 1981]. During the 1973-1976 excavations, work was confined to the lower terrace of the fortress,

150 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

on the area above the Tekhuri river. In 1977, however, the completion of a road to the top of the mountain allowed the expedition to start excavations in the citadel. Work began with the clearing of the towers in the fortification wall of fallen masonry. A number of articles in the first volume of reports from Nokalakevi-Archaeopolis are devoted to publishing the most important archaeological materials recovered between 1973-1977 [Abdushelishvili and Tsiuma 1981; Gvin- chidze 1981; Kaukhchishvili 1981; Lekvinadze and Khvedelidze 1981]. Coin finds belong almost exclusively to the late 5th-early 7th centuries AD (these are Byzantine coins of the Emperors An- astasius to Maurice 491-602AD). Metalwork found between 1973 and 1977 includes a variety of military and domestic artefacts [Zakaraia 1981]. The most interesting find was a significant por- tion of a bronze ‘choros’ (a light hanging introduced in Byzantine domed churches from the 7th century) with a cruciform personal monogram reading ‘Evstrat’, and another inscription reading ‘OEOTOKE’ (meaning Virgin), which was found in the two-storey palace. Between 1978 and 1987 work continued both in the upper citadel and on the lower terrace [Zakaraia 1987; 1993]. The remains of two churches were fully exposed in the central part of the lower town near the Forty Martyrs Church (Figure 9). The first of these was dated to the middle or the second half of the 4th century. It is rectangular in plan, with a semi-circular apse at its east end. It was replaced by a large, three-aisled basilica erected directly over it in the middle of the 5th century. This was in turn destroyed after which another aisled basilica was built to its north. This, the extant Forty Martyrs Church, appears to have undergone several phases of repair and extension during the middle ages, finally becoming a domed church [Kapanadze 1987]. Two other new buildings were discovered in the lower terrace during this period. In the south- eastern part of the site, near the fortification walls, the stone foundations were uncovered of a rectangular building, which possibly housed the military commanders of the Byzantine and Laz garrison of the 5th/6th century. The foundations of another building – a small basilica with an apse situated to the west of the Forty Martyrs church – has been interpreted as a baptistery. Work also continued in the upper citadel, including excavations in the multi-phased towers at the north- west and the east end of the fortified area atop the hill. In the southwest corner of the citadel another small gate was discovered, and clearance along the southern citadel wall provided a clearer indication of the manner in which the fortification had been laid out. In 1990, with Georgian independence from the Soviet Union on the horizon, limited archaeo- logical excavation took place at three sites. The first was located to the west of the lower town, about 50m north of the tunnel that provided secure access to the Tekhuri. Excavations at this lo- cation revealed a square building orientated northwest-southeast, the southwest wall of which was 26.5m long. The walls that run northeast from this survived to a length of 7m, were 1.2m thick and not more than 1.5m in height. The walls define an area of more than 100 cubic me- tres. The main southwest wall includes two pipes which extend the full width of the wall, and together with hydraulic mortar led to the building being interpreted as a reservoir which sup- plied water to a 5th/ 6th century AD bathhouse 70m to the south, down a steep slope. Excavation of deposits within the building produced archaeological material from various periods, some of which was present as a result of colluvial movement including Hellenistic pottery, and two sherds of pottery dated to the 9th-11th centuries AD. In the same year work took place on some of the interior fortifications located in the southeast of the lower town, near the first palace. Surviving walls were conserved to prevent their collapse,

151 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

and a small excavation associated with this work produced only two small sherds of pot and two sherds of amphorae dated to the 4th-6th centuries AD. Excavations also began to the northeast of the ‘bell tower’. As is the case with much of the lower town, particularly at its northern edge where the terrace meets the bottom of the steep slope to the north, there was a great deal of mixing of material resulting from colluvial movement. For this reason fragments of pipe-handled vessels from the 6th-4th centuries BC were found in the same contexts as Hellenistic-period beads and ceramic, and glass and metal wares dating to the Early Byzantine period. Excavations from 1973 to 1990 in the eastern and central part of the lower terrace produced finds from the main periods of occupation of Nokalakevi, namely the 8th - 7th centuries BC; the Early Antique (6th/5th centuries BC), and Hellenistic (4th-1st centuries BC) periods; and the 4th - 6th centuries AD. Finds of the 8th-7th centuries BC include a variety of precious and semi-precious stones relating to bead-manufacture; evidence of metal-working; and the discovery of a large number of fragmentary, double-headed zoomorphic figurines (Figure 5). Occupation of the site appears to have been particularly intensive during the 6th - 4th centuries BC, with ceramic finds from this period representing a range of typical domestic wares - sherds of pithoi, cooking pots, jugs, bowls, drinking vessels etc. The numerous finds of the th4 - 6th centuries AD, unearthed between 1973 and 1989, included local imitations of Roman and Byzantine red slip ware. Evi- dence of on-going occupation of the site, though on a smaller scale, was found in the form of two distinct groups of later ceramic: one group with deep dimples and incisions made in the thicker elements (for example the base and handles) before firing; the second group made of white clay, with hard, thin walls and a burnished surface. Some examples were decorated with patterns in red slip. Both groups of late pottery are dated stylistically to the 7th - 11th centuries AD [Lekvinadze 1987]. Relatively few coins were found during the excavations that took place between 1973-1989 and, apart from the famous hoard from “Schneider’s tower” (Figure 7), this is true of most of the archaeological investigations at Nokalakevi, though individual coins of Hadrian (117-138), and Constantius Chlorus (293-306) or Constantine the Great (306-337) were found. Colchian ‘tetri’ of the 4th century BC constitute the bulk of the numismatic material found at Nokalakevi. Byzantine coins - reflecting the political, economic and military relations between Lazika and Byzantium - were also found in small numbers, as were West Georgian ‘kirmaneuli’, the standard currency in the 13th to 15th centuries. There were also finds of Turkish coinage, evidence of the expansion of Ottoman influence into the area [Abramishvili 1987, 1993]. Seasons Twenty Two to Twenty Five: A difficult decade. During the difficult period that followed Georgian independence from the Soviet Union in 1991 there was little continuity, with only four seasons of excavation (between 1995 and 1998) undertaken at Nokalakevi in the years immediately following Georgian independence. These were generally small archaeological investigations undertaken by Janashia Museum staff, often unpaid and with little or no institutional support. Work in this period included continued excava- tion in the trench northeast of the bell tower. Excavation through the mixed colluvial deposits produced very interesting archaeological material - including sherds of pitchers, pots, ampho- rae, jugs and other ceramic vessels; military weapons; and a Byzantine coin. Some conservation work took place in 1994, but with a greatly reduced budget and challenging political situation, a great deal more ingenuity was required. Senior army officers from the Senaki barracks were

152 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

persuaded to supply soldiers to help with the work that was undertaken that year. In 1995 small investigations took place towards the west of the lower terrace, on agricul- tural land to the north of the expedition dig house. This part of Nokalakevi had been virtually unstudied yet, while cultivating the ground, locals had unearthed dressed limestone indicating a nearby building. Initial work involved the digging of test pits, which produced no structural evidence. As a result it was initially interpreted as a processing area for building material, how- ever the results of further investigation in 1996 added some further credence to the notion that a building was located nearby. The focus of excavation in 1996 was on the Hellenistic necropolis situated in the eastern half of the later town, and extending east of the extant walls. Work here in the early 1980s had re- vealed a significant number of graves. A single trench was opened, and no further graves were found. Being located at the bottom of the steep slope, archaeological layers were shown to be subject to the same colluvial movement that had produced very mixed upper layers in other areas of Nokalakevi. Seasons Twenty Six to Forty: The Anglo-Georgian Expedition to Nokalakevi (AGEN). By the end of the 1990s, initial discussions were already underway between the Nokalakevi team at the Janashia Museum and British specialists about establishing a collaborative expedition, and 2001 witnessed the first season of the Anglo-Georgian Expedition to Nokalakevi (kartul-ingli- suri ekspeditsia nokalakevshi). AGEN was conceived in discussions between Ian Colvin (a Byzantine historian attached to Cambridge University) and Prof. Davit Lomitashvili (then at the S. Janashia Museum and now Deputy Director of the National Agency for Cultural Heritage Preservation of Georgia). Co-founders and project directors for that first season also included Dr Besiki Lortkipan- idze (a leading Georgian historian and now curator of the Parmen Zakaraia Nokalakevi Architectur- al-Archaeological Museum-Reserve) and Nick Armour (a field archaeologist with the Cambridge Archaeological Unit). In 2002, Dr Nino Kebuladze (Head of Restoration at the S. Janashia Museum), Dr Paul Everill (a field archaeologist, now Senior Lecturer in Applied Archaeological Techniques at the University of Winchester), and Benjamin Neil (osteoarchaeologist with the Cambridge Archae- ological Unit) joined the staff, with Dr Nikoloz Murgulia (now of the S. Janashia Museum) joining the following year. Excavating every year since 2001, 2015 was the 15th season of Anglo-Georgian collaboration and the 40th season of excavation at Nokalakevi. An edited monograph, reporting the results of the first ten seasons of AGEN’s work, was published in 2014 [Everill 2014]. A short-lived Georgian-Swiss expedition excavated in Nokalakevi in 2006 and 2007, however the continuity in terms of specialists involved and the concurrence of the work means that these excavations are not considered as additional seasons. Trench A. The expedition’s first trench was opened on the 18th July 2001 and originally measured 9m x 10m, including an area to its southeast that had previously been opened along the inner wall in 1995. It was extended to its final size of 13m east-west x 13.5m north-south in 2004. The trench was orientated parallel to the fortification walls and was situated about 5m to the north of the main, eastern, gate - immediately to the west of the 6th century AD steps and their associated foundations. The 2001 season (and the extension of the trench during the 2004 season) encountered

153 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

modern overburden, containing very mixed deposits including finds from the Hellenistic period through to the 20th century. Directly underlying these deposits were large sections of masonry that had fallen from the fortification walls sealing the underlying deposits, testament to the long- term degradation of the standing Byzantine remains once they had ceased to be maintained. Underneath the masonry, deposits relating to various phases of wall construction from the 4th to the 6th centuries AD were revealed, along with associated occupation. Somewhat surprisingly, though perhaps providing evidence of ground clearance at the time of the Laz/ early Byzantine fortification of the site, Hellenistic period (4th - 1st centuries BC) burials were revealed from 2003 onwards, directly underlying the earliest of these deposits. These burials included flexed inhu- mations with associated jewellery, cremations and inhumations within amphorae and cooking vessels. Evidence suggests that the Hellenistic period was a time of great change at Nokalakevi, with structures dating to that time stratigraphically underlying the burials, though physically very close. This may indicate a shift in the focus of the settlement in those centuries and absolute dating techniques may, in the future, further clarify the temporal relationship between these phases, as further excavations elsewhere at the site shed more light on settlement foci. Early Antique period deposits (6th-4th centuries BC) were revealed from 2010, and the impression from the structural evidence for the 6th-5th/ 4th centuries BC is largely one of continuity into the Hel- lenistic period. The orientation and alignment of the walls certainly suggests that occupation of the site continued into the Hellenistic period, but more noteworthy is the notion that some of the structures, or at least phases of them, may have overlapped. The presence of a bronze scale of 6th-4th centuries BC armour and Attic pottery in one layer, at a time when the Black Sea coast was being settled by Greek traders, represents an interesting indication of the significant inland trade of Greek goods. The 8th/7th centuries BC are also well-represented in Nokalakevi, and Trench A provided further evidence for this period. It is interesting to note, in light of the discovery of an early palaeochan- nel underlying the archaeological deposits, that no structures or graves were found, but that the deposits seemed primarily related to ritual activity – in particular the significant numbers of broken double-headed zoomorphic figurines. Given that this must have been very wet ground for a considerable period after the palaeochannel had filled, with groundwater continuing to move down it as the expedition’s own recent experience in Trench A testifies, it seems entirely plausible that this was a rather liminal area in the 8th/7th centuries BC – i.e. marshy ground on the edge of the habitable area higher up the slope. It seems reasonable to suggest that these two aspects are in fact related, and that the Early Iron Age inhabitants of Nokalakevi were deliberately destroying, or sacrificing, the zoomorphic figurines prior to their deposition at the edge of, or in, the marshy ground. This kind of water-/ marsh-edge rituality is well-documented elsewhere (in- cluding examples such as the fens of East England, where traditions of ritual deposition emerge by the late Bronze Age and continue for many centuries [Rogers 2013]; and the bogs of Northern Europe) and it may well explain the pattern of archaeological evidence observed in the lower de- posits in Trench A. The Trench was finally excavated down to natural in summer 2016, with 3.5m of stratified deposits revealing the full extent of human activity in the area. Trench B. The second trench was opened on the 21st August 2002 and originally measured 7.5m x 7.5m. It was extended to its final size of 7.5m east-west x 20m north-south in 2003. It was situated 30

154 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

metres to the north east of the Forty Martyrs Church, which was first constructed in the 6th cen- tury AD and still serves as the main church for the modern settlement of Nokalakevi. Elements of the ecclesiastical precinct wall survive above ground to the south and west of the trench, along with a small square building with arches and a vaulted ceiling ten metres to the south. This would historically have served as an entrance to the precinct, and is used as a makeshift bell tower by the current church authorities. Trench B was located on the northern edge of the ‘lower town’ of historic Nokalakevi, where flatter ground gives way to the steep hillside to the north, im- mediately west of the small trench first opened in 1990. Archaeologically, Trench B was sealed by a number of thick layers representing several epi- sodes of colluvial movement. The upper layers contained a very mixed assortment of finds – plastic and metalwork dating from the second half of the 20th century; 19th century pottery and metalwork; Byzantine pottery and glass and Hellenistic pottery. Photographs of the area around the site, taken in the 1960s and 1970s, show the hillside to the north under low shrub, rather than mature trees like today. Schneider recorded the line of the fortification walls and some of his plans show the cemetery walls still clearly above ground at this point. Putting these facts together suggests that the hillside was cleared of vegetation to allow further investigation and this resulted in a substantial quantity of material from the hill being deposited around the area of Trench B in the last 70 years. If the tree-felling was for the benefit of, or resulting from the work of, the first archaeologist to excavate in Nokalakevi, it is ironic that these actions appear to have ultimately led to the burial of the walls and the creation of a new archaeological record. Its areas of principal archaeological interest can be summarised as follows: 1. A Christian cemetery first established in the early Byzantine period (about the 5th/6th cen- tury AD), revealed in the southwest corner of the trench. This produced 37 burials in excavations from 2002-2005 and 2009-2012, with two further partial burials revealed beyond the crude wall which apparently defines the cemetery enclosure. This wall measures between 0.8m and 0.92m wide, with a maximum surviving height of 1m. It was constructed from a mixture of riverstone; large, natural limestone boulders; and small, dressed limestone blocks, possibly robbed from the first two phases of fortification wall (4th–5th century AD). It also includes one large, dressed limestone block, presumably robbed from the final, early Byzantine (6th century AD) phase of for- tifications, as a cornerstone at the outside northeast corner. This indicates that the cemetery wall must certainly post-date the late-6th century, though it is not known precisely when the fortifica- tions ceased to be maintained. The requirement to preserve the wall precluded the archaeologi- cal assessment of directly underlying layers, but it seems likely that it was constructed no earlier than the 17th century, possibly replacing an earlier boundary. The intercutting nature of many of the burials, combined with associated 5th/6th century AD material culture in some graves, and the apparent survival of coffin timber in one other appear to suggest that the cemetery was in use from the early Byzantine period through to the 20th century - with the intensity of its use re- flecting the periods of growth and decline of the settlement. All bar three burials appear to have been laid out in a standard Christian manner, with the three – two adults and a neonate – being orientated north-south (feet to the south) in a supine position, and may represent a socially lim- inal family group buried at the northeast corner of the cemetery. 2. A Hellenistic period clay and timber structure and associated deposits located towards the north of the trench. Somewhat surprisingly these remains were sealed by a series of substan-

155 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

tial colluvial layers. This underscores the oddity of the apparent absence of layers from the first three centuries AD at the site, and perhaps also indicates that settlement in the early Byzantine period was not as dense as might have been expected. However, a series of absolute dates for the sequence, to be obtained during future investigations, will allow for more detail to be ap- plied in future analyses. The method of construction can be inferred from the small area of the building exposed, and is consistent with the evidence of analogous buildings revealed from 2006 onwards in Trench A. It involved the laying of an unbonded line of large limestone blocks as a foundation, or sill, onto which was placed a wooden beam or beams. Upright posts measuring approximately 0.1m in diameter, such as the ones found as charcoal, or as impressions in pieces of daub, were fixed to this horizontal beam. A wattle and daub wall was constructed on this framework. There was no archaeological evidence for the roofing material, which is most likely to have been wooden shingles or thatch. To the south of the building a large yard surface, formed of a dense layer of angular limestone pebbles and cobbles, was found. The cemetery area was excavated down to natural deposits in 2012, with the area north of the cemetery wall excavated from 2003-2005; and then from 2010-2014. Trench B was finally com- pleted towards the start of the 2014 season. This was the first of the recent trenches to expose natural deposits, which, in this part of the lower terrace, consisted of a very firm reddish brown clay. The first archaeological deposit overlying this was colluvial in character, with a substantial quantity of angular limestone cobbles. The material culture retrieved from it included worked flint, the butt end of a polished stone tool with a drilled hole for the haft, and pottery that was OSL-dated to the Bronze Age. Trench C. Following the completion of Trench B, Trench C was opened on the 3rd July 2014 and mea- sured 5m x 5m. It was situated 9.5 metres to the west of the current Dig House, which was first constructed as the Director’s house around 1977. Elements of the original Dig House, such as concrete steps and the north wall of the ground floor where it also served as revetment against the slope, survived to the east and north of the trench. This building was originally constructed as the village hospital towards the end of the 19th century, or early in the 20th century, and ap- pears in photographs taken by D. Gozalishvili during the 1930-31 expedition. The excavation of Trench C allowed for the first investigation of the western end of the ‘lower town’ with modern techniques. The results were not dissimilar to those observed in Trench B and are revealing as much for what is absent as what its present. Aside from the fascinating oppor- tunity to excavate the material remains of the expedition’s Soviet-period precursor, those upper layers underlying the dig house were colluvial sediments, containing mixed material culture, including OSL dated ceramics from the Hellenistic to perhaps as late as the 12th century AD. As was the case in the north part of Trench B, the first in situ ancient remains, observed in 2015, were a wall sill/ base formed of unbonded limestone blocks. Underlying this, with further parallels to Trench B, was a minimum of one metre of colluvial sediments overlying a primary archaeological layer containing ceramic, OSL-dated to the Bronze Age, but no related structural evidence. The trench was completed in the last few days of the 2015 season. Trench D. The expedition’s fourth area of activity involved archaeological work in and around the Forty Martyrs’ Church, which took place from 25th August to the 12th September 2014. It followed a

156 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

request from Bishop Shio (Mujiri) of the Senaki and Chkhorotsqu Diocese of the Patriarchate of Georgia for the investigation of certain elements of the site, and the work was undertaken by a small team after the main field season had been completed. The work was funded by the Na- tional Agency for Cultural Heritage Preservation of Georgia, and undertaken by members of the Anglo-Georgian Expedition to Nokalakevi utilising the expedition’s established excavation and recording methodology. Trench D in fact consisted of two foci (the south nave and the south porch) which were unified into one area of investigation during the excavation. Concurrent with the excavation inside the church, an examination of the Dadiani family crypt against the external north wall was also undertaken to assess condition. The archaeological work within the church was primarily intended to evaluate the survival of evidence relating to the Church’s dedication to the Forty Martyrs and revealed a grave in the south porch which had been capped by rows of round pilae, presumably reused from the an- cient bathhouse. Although the pilae had been disturbed by the digging of a small modern pit, perhaps for the purpose of concealing a silver icon frame from the Bolsheviks, it seemed likely that they originally consisted of four rows of ten pilae referencing the Forty Martyrs. The porch had been thought to date to the 16th or 17th centuries, however the skeleton was associated with a distinct type of buckle. The closest Georgian analogy was with buckles discovered in Samtavro, which are similar to Avarian types and are dated to the 7th century AD. The buckle found in No- kalakevi was even more similar to the buckles discovered in the Volga Federal District, in the cemetery of the earlier Bulgarians, and these are dated to between the 8th century and the 10th century AD. Consequently the burial has been dated broadly to the 7th-10th centuries AD and, while the porch may have been constructed over an earlier tomb, it seems more likely that it was integral to this structure and that the porch itself is far earlier than previously thought. Trench E. The fifth trench was first opened in 2015 and measured 10m north-south x 9m east-west. The trench was orientated parallel to, and 30m from, the eastern fortifications and was therefore not on a true north-south alignment. An open area trench was opened at this location to investigate properly the results of a small test trench that had been excavated outside the walls in 2006. This original trench had revealed archaeological layers indicating the presence of a significant defensive ditch, but the small size of the trench itself made it impossible to draw conclusions. At the time of writing work is ongoing in Trench E. Trench F. The most recent trench was opened at the start of the 41st season, in 2016. It was located at the northern edge of the old Trench B, with the express purpose of properly investigating the Hellenistic period structure observed there in 2005. Initially measuring 10m east-west x 5m north-south, Trench F revealed a continuation of the east-west wall line as well as further walls north of the original Trench B, again indicated by lines of unbonded limestone blocks. Further work is required in order to fully understand the various wall alignments, however early indica- tions are of a series of overlying phases of construction and occupation at the base of the slope, above the lower terrace. Conclusions The site of Nokakalevi has been well-studied since the first archaeological investigations in 1930, and yet the size and complexity of the site means that a great deal more needs to be done.

157 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Looking at the results of the significant excavations from 1973 onwards, the story of Nokalakevi appears to be one that begins in the Bronze Age with limited human activity but, as yet, no evi- dence for actual settlement of the site. The first indication of more significant human presence appears in the 8th/ 7th centuries BC, with concerted ritual activity. This included, most notably, the deliberate breaking and deposition of double-headed zoomorphic figurines at the edge of the habitable area where it bordered a waterlogged, possibly marshy, space on the eastern lower terrace. However, to date no structures have been identified from this period and it is possible that this evidence will be found further up the slope. From the 6th century BC, possibly reflect- ing more favourable climatic conditions that also saw the flourishing of the Kingdom of Colchis, there is ample evidence of settlement on Nokalakevi’s lower terrace, and this continues through the Hellenistic period. There is sparse evidence of activity at the site in the first three centuries AD, before it became an important regional centre in the 4th century. For the following three centuries it was a key military fortification of the Kingdom of Lazika, with the final phase of walls and towers being constructed to accommodate a combined Laz and Byzantine garrison in the 6th century AD. It is unclear exactly when the fortifications slipped into disrepair. They may have been slighted during Byzantine-Persian warfare at the beginning of the 7th century or – accord- ing to Georgian historical sources – by Arab invaders in the early 8th century. Archaeological work at Nokalakevi will continue, as the current expedition and our successors shed more light on the fascinating history of the site. As methods evolve, and scientific tech- niques become more precise, we can only imagine what details might emerge over the next 40 seasons of excavation.

Acknowledgements. The authors wish to gratefully acknowledge the important contribution made to our under- standing of the site by those specialists who came before us. We would also like to extend our thanks to all those who have participated in excavations at Nokalakevi, particularly to the more than 200 students from Georgia, Britain and elsewhere who have taken part in the Anglo-Geor- gian Expedition since 2001. AGEN is especially grateful to the British Institute at Ankara, and the British Academy Black Sea Initiative, which helped fund our early seasons.

158 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Bibliography:

Abdushelishvili, M., Tsiuma, N. 1981: ‘Anthropological materials of Hellenistic times from Nokalakevi’, in P. Zakaraia (ed.), No- kalakevi-Archaeopolis I. Archaeological excavations 1973-1977. Tbilisi, 200-219 (in Georgian). Abramishvili, T. 1987: ‘Coins from Nokalakevi and Nodjikhevi’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis II. Archaelogical exca- vations 1978-1982. Tbilisi, 274-287 (in Georgian). Abramishvili, T. 1993: ‘The Numismatic Material found at Nokalakevi’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis III. Archaeo- logical excavations 1983-1989. Tbilisi, 270-272 (in Georgian). Arnold, C. 1999: Christian Antiquity and Roman Catholic German Science. The Freiburg Church Historian and Archaeologist Jo- seph Sauer (1872-1949). Paper presented at the Thirteenth International Conference on Patristic Studies, Oxford, 16.-21. August 1999. https://www.academia. edu/11268598/Christian_Antiquity_and_Roman_Catholic_German_Science._The_Freiburg_ Church_Historian_and_Archaeologist_Joseph_Sauer_18721949_Thirteenth_International_Conference_on_Patristic_Stud- ies_Oxford_16.-21._August_1999_ (Accessed December 2016). Colvin, I., Lortkipanidze, B., Murgulia, N. 2014: ‘Historical overview of Colchis-Egrisi-Lazika’, in P. Everill (ed.), Nokalakevi – Tsikhegoji – Archaeopolis. Archaeological excavations 2001-2010. Anglo-Georgian Expedition to Nokalakevi. BAR International Se- ries 2612. Oxford, Archaeopress. Dubois de Montpéreux, F. 1839: Voyage autour du Caucase, chez les Tcherkesses et les Abkhases, en Colchide, en Géorgie, en Armé- nie et en Crimée. Volume III. Paris, Librairie de Gide. Everill, P., Marter, P., Lomitashvili, D., Murgulia, N. 2011: ‘Mapping Archaeopolis: GPS survey at the multi-period site of Nokal- akevi’. Bulletin of the Georgian National Museum. Series of Social Sciences #2 (47-B): 117-130 (in Georgian). Everill, P. (ed.) 2014: Nokalakevi – Tsikhegoji – Archaeopolis. Archaeological excavations 2001-2010. Anglo-Georgian Expedition to Nokalakevi. BAR International Series 2612. Oxford, Archaeopress. Gvinchidze, G. 1981: ‘Burials at Nokalakevi’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis I. Archaeological excavations 1973-1977. Tbilisi, Metsniereba: 150-183 (in Georgian). Gvinchidze, G. 1988: The history of the population of Colchis in the Hellenistic period. Tbilisi, Metsniereba (in Georgian). Gozalishvili, G. 1981: ‘Several words about the beginning of Archaeological excavation in Nokalakevi’, in P. Zakaraia (ed.), Noka- lakevi-Archaeopolis I. Archaeological excavations 1973-1977. Tbilisi, Metsniereba: 243-248 (in Georgian). Grigolia, G., Okropiridze, N., Japaridze, V., Tsirghvava, V., Chartolani, Sh. 1972: ‘The results of the West Georgian Exploratory Archaeological Expedition work in 1971’. Archaeological investigation in Georgia in 1972: 29-31 (in Georgian). Grigolia G., Pkhakadze, G., Baramidze, M., Lortkipanidze, G. 1973: The results of the West Georgian Exploratory Archaeological Expedition work in 1968. MSKA. Vol. 5: 17-37 (in Georgian). Kapanadze, T. 1987: ‘Basilicas at Nokalakevi’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis II. Archaeological excavations 1978- 1982. Tbilisi, Metsniereba:: 90-125 (in Georgian). Kaukhchishvili, T. 1981: ‘Greek Inscriptions from Nokalakevi’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis I. Archaeological exca- vations 1973-1977. Tbilisi, Metsniereba: 197-199 (in Georgian). Lekvinadze, V. 1987: ‘Mass Archaological Material of the 4th to the 11th centuries discovered on the site of Nokalakevi in 1978-1982’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis II. Archaeological excavations 1978-1982. Tbilisi, Metsniereba: 237-248 (in Georgian). Lekvinadze, V., Khvedelidze, L. 1981: ‘Mass archaeological finds from excavations at Archaeopolis’, in P. Zakaraia (ed.), Nokal- akevi-Archaeopolis I. Archaeological excavations 1973-1977. Tbilisi, Metsniereba: 120-149 (in Georgian). Murgulia, N. 2013: ‘The fortification system of the kingdom of Egrisi in the 4th-6th centuries’. Unpublished doctoral thesis. Tb., (in Georgian). Muskhelishvili L. 1987: ‘The Journal and results of the excavations 1930-1931’, in P. Zakaraia (ed.), Nokalakevi-Archaeopolis II.

159 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Archaeological excavations 1978-1982. Tbilisi, 289-305 (in Georgian). Rogers, A. 2013: ‘Water and Roman Urbanism: Towns, waterscapes, land transformation and experience in Roman Britain’. Leiden, Brill. Schneider, A.M. 1931: ‘Archaeopolis (Nokalakewi)’ Forschungen und Fortschritte 27: 354-355. Serena, C. 2015: Excursions in the Caucasus: From the Black Sea to the Caspian Sea 1875-1881 (Edited by P.F. Skinner). New York, Narikala Publications. Zakaraia, P. (ed.) 1981: Nokalakevi-Archaeopolis I. Archaeological excavations 1973-1977. Tbilisi, Metsniereba (in Georgian). Zakaraia, P. (ed.) 1987: Nokalakevi-Archaeopolis II. Archaeological excavations 1978-1982. Tbilisi, Metsniereba (in Georgian). Zakaraia, P. (ed.) 1993: Nokalakevi-Archaeopolis III. Archaeological excavations 1983-1989. Tbilisi, Metsniereba (in Georgian).

Figure Caption:

Figure 1: Regions of Georgia, showing Nokalakevi in Samegrelo, northeast of Senaki. Figure 2: 3D model of Nokalakevi (looking north) derived from GPS survey in 2009 (Everill et al 2011). Figure 3: Drone photo of Nokalakevi, looking east (© National Agency for Cultural Heritage Preservation of Georgia). Figure 4: Annotated site plan of Nokalakevi (co-ordinates in WGS84 UTM Zone 38N). Figure 5: An example of the double-headed zoomorphic figurines from Nokalakevi (Everill 2014:xiii). Figure 6: Ruins of Nokalakevi around the Forty Martyrs’ Church – Drawing by P. Sellier, after Dubois de Montpéreux (Serena 2015: 23). Figure 7: Byzantine coin hoard discovered during Schneider’s excavations at Nokalakevi (Everill 2014). Figure 8: The eastern walls of Nokalakevi in the early stages of cleaning and conservation (Zakaraia). Figure 9: Drone photo showing the extant Forty Martyrs’ Church, left, the foundations of two 4th/ 5th century churches, centre, and the remains of the ‘palace’, on the right (©National Agency for Cultural Heritage Preservation of Georgia). Figure 10: Anglo-Georgian Expedition to Nokalakevi, location of Trenches A to E.

160 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

P. EVERILL, D. LOMITASHVILI...

Figure 1

Figure 2

161 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Figure 3

Figure 4

162 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Figure 5

Figure 6

163 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Figure 7

Figure 8

164 This article has been published by the Georgian National Museum in Iberia-Colchis, available online at http://dspace.nplg.gov.ge/bitstream/1234/242318/1/Iberia_Kolxeti_2017_N13.pdf. Copyright © 2017, Georgian National Museum.

Figure 9

Figure 10

165